[CSSWG] Minutes Telecon 2014-11-12

CSS3 UI
-------

  - This discussion focused on two different items: the previous
      decision of the working group to add Florian as a co-editor to
      CSS3 UI and how CSS3 UI should progress forward.
  - In regards to co-editorship, it was decided that Florian should
      function as co-editor for CSS UI Level 4 instead of for level 3.
  - Progress on CSS3 UI will hopefully move forward more quickly now
      that tantek has time to focus on it and more unstable items
      will be moved back to level 4.
  - The work Florian already has done to take previous commits into
      bikeshed will still be handled by the two of them and they
      will work together to get the edits Florian found into CSS3 UI.
  - RESOLVED: Add Florian as a CSS UI level 4 editor, not to level 3
      and they will work together on publishing the pieces that
      Florian already prepared for level 3

Exclusions and Position
-----------------------

  - RESOLVED: New WD for Exclusions and Position

pseudo-elements
---------------

  - The FPWD of pseudo-elements will happen after language is added
      for ::grammar-error and ::spelling-error, including
      information pertaining to the security concerns.

Restructuring CSS Generated Content
-----------------------------------

  - There was a lot of support for the general path dauwhe is taking
      on Generated Content and it was agreed that as issues come up
      they'll be discussed further on the list.
  - RESOLVED: Proceed as outlined by dauwhe for Generated Content
      and continue conversing on the ML

Host for 'content' property
---------------------------

  - The 'content' property should stay where it is.

:hover pseudo
-------------

  - The language was wrong and will be changed

Where does ::selection inherit from?
------------------------------------

  - fantasai asked for responses on the mailing list about where
      ::selection should inherit from.  The thread is available
      here: http://lists.w3.org/Archives/Public/www-style/2014Nov/0188.html

===== FULL MINUTES BELOW ======

Present:
  David Baron
  Bert Bos
  Bo Campbell
  Tantek Çelik
  Dave Cramer
  Alex Critchfield
  Elika Etemad
  Simon Fraser
  Sylvain Galineau
  Daniel Glazman
  Dael Jackson
  Chris Lilley
  Peter Linss
  Shinyu Murakami
  Keshav Puttaswamy
  Florian Rivoal
  Simon Sapin
  Dirk Schulze
  Alan Stearns
  Lea Verou
  Greg Whitworth
  Steve Zilles

Regrets:
  Rossen Atanassov
  Mike Miller
  Simon Pieters
  Anton Prowse

  Scribe: dael

  glazou: Let's start.
  glazou: Florian, I got your request late.
  <Florian> http://www.w3.org/mid/8ADFE4C9-579D-4DF2-B6CD-53FF9C2D6ADB@rivoal.net
  <Florian> http://www.w3.org/mid/FD5224A3-309C-4B65-BD71-D50D4F0314BC@rivoal.net
  Florian: I got two points, they're linked above.

  glazou: Anything else? Other extra items?
  Bo_Campbell: I wanted to introduce an item if we have time.
  glazou: I missed your e-mail announcing you joined the group. We
          usually do the introductions on the ML.
  Bo_Campbell: I did send an introduction, but I'll do that again.

CSS3 UI
-------

  <smfr> http://dev.w3.org/csswg/css-ui/
  <tantek> dael, see my email to www-style on this
           http://lists.w3.org/Archives/Public/www-style/2014Nov/0199.html
  tantak: I requested we talk about what changes first.
  glazou: We want to revisit the discussion.
  tantek: I wanted to do edits separately of editorship.

  glazou: The context is in Sophia we decided that CSS3-UI was long
          unattended. We don't have a good track record in general
          of PRs and REC. We retired a few documents as notes and
          gutted notes, the rest we wanted active editorship.
  glazou: CSS3-UI was mentioned in Sophia with no complete decision.
          It was 2 months before TPAC. So at TPAC we decided not to
          remove your editorship, tantek. The WG only decided to add
          another name, someone who is more available to do
          editorial stuff to make progress. That's all I have to say
          as the chair. I'll participate as a member.
  tantek: That's the first I've heard about this. The first I heard
          about the new heartbeat requirement was from plinss at
          TPAC. My understanding is that the group is now, as you
          said, making a priority of end of life-ing a document or
          make it more actively published. Is that correct?

  glazou: That plus there's new suggestions coming, for instance
          from Bloomberg, and that was discussed in Sophia.
  tantek: Was there resolution to add something? What was resolved
          to CSSUI? The previous was to keep the working items in 3
          and postpone the rest.
  glazou: There were a few additions.
  Florian: It was recognizing and accepting items to work on.
  tantek: That makes sense to accept things to work on. There was a
          drive to wrap up CSS3-UI as is and put anything squishy
          and uncertain in 4. I'm fine with caret-color, but I
          think, being that items, are beyond the scope of that
          plan, I'd push against them being added to level 3.
  glazou: I think it depends on implementations, so it's not in our
          hands.
  tantek: I agree on that.
  <astearns> (some of the) CSS UI resolutions from Sophia:
             http://lists.w3.org/Archives/Public/www-style/2014Oct/0260.html
  * astearns notes that we also resolved to add Tab as a co-editor
             to CSS UI in Sophia:
             http://lists.w3.org/Archives/Public/www-style/2014Oct/0295.html
  * sgalineau thinks the issue of how we make the spec progress is
              orthogonal to whomever edits it

  glazou: I would like to understand, you object to a new co-editor,
          not a new editor.
  glazou: This is the first time in the history of the WG, so I'd
          like to understand why.
  tantek: I objected to the order of discussions. I'd object to this
          about any document. My understanding of the way the WG has
          worked is there is an order of operations where if the WG
          decides we need a new draft or changes, the WG can decide
          even without the editor there and give a certain date for
          completion. My understanding this has been at least 2
          weeks
  tantek: I object to skipping completely and changing editors.
  glazou: It happened with howcome.
  fantasai: Your understanding of how the WG has worked is wrong.
            I've been here for 10 years and we've added co-editors
            to docs and there's never been a requirement or
            consideration of competence for the old editors. What we
            have considered is if the person is willing and
            interested. If there was a case of incompetence your
            process makes sense. We have had people leave and
            disappear.
  tantek: We have a different memory than.
  fantasai: I can read off minutes.
  * sgalineau is not sure the way the WG has done things in the past
              is an interesting or relevant topic
  <SimonSapin> I don’t think we asked Melinda Grant or Håkon before
               adding me as co-editor to css-page
  tantek: This WG has had a healthier interaction than you describe.
          I don't want to see that go forward. We're talking about a
          way forward. I also object to brining in incompetence to
          the conversation.
  <fantasai> tantek: I don't understand why you think this is
             unhealthy. Person volunteers to help me edit my spec. I
             don't have a problem with it. WG agrees to add the
             editor. Done.
  <tantek> fantasai - rudeness is unhealthy.
  <tantek> I don't understand why you're defending that.
  <fantasai> tantek: What is rude about the process I just
             described? Nothing.
  <fantasai> tantek: What's rude is maybe that you were not present
             at the last discussion about editorship of css3-ui
  <fantasai> tantek: Yes, I agree that's rude
  <fantasai> tantek: We should not do that.
  <fantasai> tantek: but that's a different issue as to whether we
             need to prove that you cannot handle editing the spec
             before we consider volunteers to co-edit
  <sgalineau> fantasai: I think it's rude if there is no follow up
              and the editor finds out weeks later by accident.
  <fantasai> sgalineau: I'm not contesting that.
  <fantasai> sgalineau: Is it rude if the editor was present in the
             original discussion?
  <sgalineau> fantasai: just saying it's not necessarily rude to
              make those calls without the editor.
  <sgalineau> fantasai: I think we're off in the weeds and spending
              way too much time discussing everything except what
              the WG wants to do with the spec

  Florian: I think the example of howcome wasn't interesting here. I
           think another would be when I was editing MQ and I was
           doing it moderately actively. TabAtkins had an interest
           in working on it, he was added, did edits, and there was
           no consideration to not responding. TabAtkins worked as a
           co-editor and responded to requests from the WG and made
           edits. I think that was useful.
  Florian: At the moment I have time and interest for the spec and
           adding my time I hope would help it move forward. This
           isn't about you responding to the group. To drive things
           forward isn't just about responding, it's about
           considering what has been added in other places and
           pushing the group. We're not retiring you because you're
           not responsive.
  tantek: Florian I'm not blaming you. I appreciate you have time
          and interest to work on CSS UI. As I've communicated
          privately, I think the more interesting work is level 4
          and that's an area that I explicitly recommend needs more
          work and could use the help of a co-editor in more ways
          that CSS3 UI
  tantek: So let's do that for 4. For 3 let's trim things in a way I
          think I can be more effective at as a single editor. I
          don't want to squash enthusiasm and want to see you help
          work on level 4 where there's renewed interest. That's how
          I'd like to move forward with you, Florian if that's okay.
          I'd rather collaborative decisions instead of overriding
          opinions.
  glazou: You're proposing a co-editor rule.
  tantek: I'm proposing for level 4.
  glazou: It has two levels to work on.
  <astearns> I think it's a fine compromise to add Florian as a co-
             editor to level 4 and have Tantek move level 3 forward

  glazou: To go back to fantasai point we have plenty of examples
          about adding editors. The last document you edited in this
          group was fullscreen. Okay, you attended a few calls, but
          rarely the full thing. You're far less involved than you
          used to be.
  tantek: I think you're changing the topic and it's not relevant.
  glazou: I think it is.
  tantek: If Florian and I can work out how we think CSS UI should
          work, I think we should be able to do that.
  glazou: The WG should decide.
  tantek: I don't think they should micromanage.
  glazou: The group decides. We're lax, but everything we do is
          decided by the WG. All our rules are decided by the WG.
          You know that.
  tantek: Rarely does the WG decide against what an editor is trying
          to do. I understand we're run by consensus, but we try and
          work with the desires of those working.
  glazou: We had consensus and we're still discussing it.
  tantek: No one is arguing about able, but I think there's a better
          way forward.
  glazou: We heard you, I'd like to hear from other WG members.

  Florian: With the level 4 proposal, I'm happy to do that. In level
           3 there's one or two things that concern me a bit. To be
           able to work on 4, it's not productive if 3 isn't
           somewhere between PR and REC.
  Florian: I currently intend to go line by line to see if there are
           things that need to be addressed, go through the wiki,
           bring things to the ML. I can forward all the resolutions
           and proposals to you, but having you as a gatekeeper
           sounds like unnecessary overheard.
  tantek: I'm not talking about delays, I'm making this a priority
          and I'm committing to working actively to dealing with the
          issues regarding CSS3 UI. I think we can work together to
          conquer the items and divide them between 3 and 4. I don't
          think this is gate-keeping, I'd like to try the way I
          proposed for a couple weeks to see how it goes. If you
          want to start 4 from the current state of 3 that's
          reasonable, but I'm more of the opinion I'd like the
          changes...
  tantek: to 3 to be more conservative. Even if they're editorial, I
          think it slows things down. I think the two levels should
          be handled differently. I'd like you to take charge of the
          much more free to make changes part where the gate-keeping
          is getting 3 more quickly to CR.
  <TabAtkins> FWIW, I'm still of the opinion that we should only
              have one active level of a module at a time, with
              "higher" modules being maintained as deltas except
              when published. Don't think we have a single instance
              of co-active module levels that were *actually*
              properly maintained, with fixes propagating both ways.
  glazou: It seems that you're the author, owner, and decider, not
          the editor. The WG is the owner and they will decide how
          to handle it.
  tantek: I think I'm looking for collaboration. I'd like to see
          Florian and I make progress.
  <bkardell> "Try it for a few weeks" seems relatively harmless...
             Why not give it a shot with review in 2 or 3 weeks
  <glazou> bkardell: yes

  Florian: I'd like to clarify when I said line by line, I don't
           mean to change things everywhere, I mean to refine things
           that are insufficiently explained or see when things have
           been more accurately described in another place.
  Florian: For example, the introduction is a bit out of date. For
           the actual text, I don't plan on making major changes and
           agree with you about it being conservative. If you have
           time to actively work great, but I'm concerned that I
           have spec changes I want to put in and sending you
           patches isn't as straight forward as doing it.
  tantek: Yes, I have time to work on this and am making this a
          priority and I recognize that I haven't in the past. So
          I'd like to work on moving this forward together.

  glazou: I'd like to hear from others.
  <tantek> I think astearns had an opinion
  andreyr: I think that Florian should be a co-editor as we voted. I
           don't see a problem with that.
  glazou: Other opinions?
  astearns: I can live with tantek's plan, but I agree with Florian
            it's unnecessarily bureaucratic so I'd prefer Florian as
            a co-editor.
  fantasai: I don't see a problem with Florian being the co-editor.
            I think since he has patches it seems silly to have him
            put it through the mailing list. I think they should be
            able to work together.
  tantek: That's not what's being proposed. I'm proposing something
          much more interactive. A two week wait is ridiculous.

  SteveZ: It seemed to me that tantek and Florian were working out a
          relationship and it seems to me that the group doesn't
          need to micromanage. I'm coming down that Florian should
          be an editor and we leave it to them to work out how they
          divide the work on 3 and 4. I know that's not exactly what
          tantek is asking, but I think that they can work it out.
  Florian: I do think it would be simpler if I'm a coeditor. I don't
           object to having tantek be the one doing it and me
           forwarding requests, but it seems unnecessary. If that's
           the only way forward, it's better than none, but it's
           process heavy.
  tantek: I don't want to be process heavy. I think dividing the
          work between 3 and 4 would be more efficient. I don't want
          to be a gatekeeper, I'll keep an eye on what you do to 4
          and I'll be very conservative on 3 and we can interact on
          IRC or e-mail or whatever. This isn't about process, this
          is about being efficient.

  fantasai: TabAtkins said earlier that 2 specs that are duplicating
            the same content doesn't work well. When there's two
            levels we have the second be a delta until the one above
            stabilizes. Florian doing level 4 means he won't be able
            to work on much.
  tantek: I don't think so. 3 is pretty stable.
  glazou: What if the WG decides in 2 weeks to add something the
          CSS3 UI. You said you want conservative. We can change our
          mind and decide to add something, will you let it be.
  tantek: Yes, if the WG decides to change we'll do it, but I think
          CSS3 UI is a release branch and active work will go into 4.
          Back-porting bug fixes will need to be work on CSS3 UI.
          That's part of the natural course of working on specs.
  <dbaron> I don't think that addresses how the work will happen to
           define the features in level 3 more carefully and
           precisely.

  glazou: So we need to move on. tantek, Florian what's the plan?
          You guys start working together and reach a gentlemen's
          agreement on how to proceed?
  Florian: I would like to point out that I'd agree more with tantek
           if the spec was in late CR.
  tantek: The reason it's not is it already has bounced a lot and I
          spent time working on the new process. I understand that
          the WG is now on the new process.
  Florian: We move them at the first opportunity.
  tantek: Okay. That's why I wanted to focus on the new process, but
          I did that too long. I'm sorry. I'm back.

  glazou: dbaron is on the queue.
  dbaron: We should wrap up.

  Florian: This is more controversial than I thought. I prepared
           commits to bikeshed the spec. I'd like to push that
           through and then we work as you described onward.
  tantek: I'm okay with that.
  <tantek> thank you Florian
  <tantek> and thank you glazou for permitting an agreement between
           Florian and myself
  glazou: Okay, we have an agreement.

  fantasai: Do we have a resolution on the editorship of CSS3 UI?
  <tantek> let Florian talk
  <tantek> please fantasai
  fantasai: Is Florian added?
  glazou: Not as an editor. They will discuss how they will work
          together.
  <tantek> Florian go ahead
  Florian: We have discussed.
  Florian: I am added as a level 4 editor, not to level 3. tantek
           will work on that and I'll forward what's needed to him.
           I'll publish what I've already prepared for 3.
  Florian: tantek, is that a correct desc?
  tantek: I agree and further commit to reducing any overhead in
          moving forward.

  RESOLVED: Add Florian as a CSS UI level 4 editor, not to level 3
            and they will work together on publishing the pieces
            that Florian already prepared for level 3

  glazou: tantek, does that resolution work for you?
  tantek: And you wanted to record the changes to level 3? Oh, yes,
          okay.

Exclusions and Position
-----------------------

  gregwhitworth: That item was from rossen and he converted them to
                 bikeshed and made editorial changes. That's what I
                 know. If you have comments, let me know, but he
                 wanted to get it on the agenda.
  gregwhitworth: That's it.
  glazou: Comments?
  <andreyr> I am in favor
  astearns: I'm in favor
  <Florian> +1
  glazou: Me too.
  glazou: Objections?
  <dbaron> what's in this new draft?

  RESOLVED: New WD for Exclusions and Position

  <dbaron> https://lists.w3.org/Archives/Member/w3c-css-wg/2014OctDec/0104.html
           says that the edits to css-position are migration to
           bikeshed and editorial fixes (typos and broken links)
  <dbaron> https://lists.w3.org/Archives/Member/w3c-css-wg/2014OctDec/0090.html
           says that the edits to css-exclusions are migration to
           bikeshed and editorial fixes (mostly typos and broken
           links)

pseudo-elements
---------------

  <fantasai> http://lists.w3.org/Archives/Public/www-style/2014Nov/0188.html
  fantasai: There's some issues. Adding ::grammar-error and
            ::spelling-error was discussed, but there was no
            resolution.
  dbaron: Do we want these specifically, or something more generic?
  <dbaron> ... for named selections

  fantasai: I think we discussed that there would be more generic,
            but these two would need to be hard coded. We could
            share syntax with functional notation, but it makes
            sense to put these by themselves, like ::selection. I'm
            open to other ideas.

  Florian: What is the current question being asked?
  <tantek> I'm confused too
  glazou: A FPWD for psuedo-elements.
  <fantasai> Question is: Do we add ::grammar-error and
             ::spelling-error
  <andreyr> I vote yes.

  fantasai: Right now about ::grammar-error and ::spelling-error
  Florian: My answer to that is yes if it includes security.
  glazou: Do we need it for FPWD?
  Florian: It's better to have it if we know it's wanted.

  gregwhitworth: What if it was more generic? A way to just control
                 the color of the squiggle, no matter how it's used.
                 I can think of other ways to use it.
  fantasai: You can style squiggle with text decoration.

  <tantek> who is editing [css-pseudo] ?
  * fantasai is
  <Florian> tantek: editors are fantasai, glazou, astearns.
            Theoretically TabAtkins as well, but he's not acted on
            it.
  <tantek> Florian thanks.
  <glazou> tantek: original authors are astearns and myself

  Florian: So, I would say it should be in FPWD since we roughly
           know how to spec it and it's good to have it there.
  <Bert> I agree with Florian: a note about security would be good
         in the FPWD.
  <tantek> I also agree with Florian re: a note about security would
           be good in the FPWD
  andreyr: I concur.
  glazou: So it seems we won't publish FPWD right now.
  fantasai: Yeah.
  glazou: So no FPWD until we add the security thing for
          ::spelling-error and ::grammar-error

Restructuring CSS Generated Content
-----------------------------------

  <glazou> http://lists.w3.org/Archives/Public/www-style/2014Nov/0126.html
  dauwhe: It has a WD from 2003. We resolved in Sophia to move some
          of GCPM here and resurrect it and I want general
          permission to apply the last 11 years of history to this
          draft. For example, counters and numbering have their own
          spec.
  dauwhe: I had contacted the old editors and they're willing to
          become former editors. I'm bikeshedding and getting it
          closer to current, so I want to make sure there's no
          objections to this large scale tearing up of the spec.

  <ChrisL> +1
  <SimonSapin> +1
  Florian: No objections. I think it's a great thing.
  <tantek> +1 and appreciate that dauwhe explicitly contacted former
           editors
  <tantek> thank you dauwhe for that
  <fantasai> dauwhe++

  dauwhe: Since this has the content property itself we'll need to
          deal with related issues here eventually.

  glazou: So objections to the proposed plan?
  glazou: I think it makes a lot of sense.

  Florian: Do we need to go through the A-H list on the e-mail?
  dauwhe: I think they're mostly fine. A lot of these are obvious.
          If small scale issues appear we can discuss on calls or
          list as needed.
  glazou: I think E & F need resolutions.
  dauwhe: I'm not ready for substantive conversations on those. I
          wanted them in for the awareness.

  <tantek> dauwhe: btw re: http://dev.w3.org/csswg/css-content/#content
           specifically 'icon' value - no one has implemented it
  <tantek> dauwhe: I had initially proposed that ('icon' value, and
           icon property), and it was in CSS3-UI, but am dropping
           from there, and am ok with you keeping it in css-content
           if you wish, but also ok with you dropping it. Up to you.

  glazou: So obj to the global plan?
  fantasai: I think it's great.
  <Bert> (I like Dave's plan)
  <andreyr> no objection
  Florian: I object to NOT doing it.

  ChrisL: Given that this is so old, it should be a FPWD as far as
          patent policy.

  RESOLVED: Proceed as outlined by dauwhe for Generated Content and
            continue conversing on the ML

Host for 'content' property
---------------------------

  astearns: It sounds like I was wrong and we plan on having it
            apply outside pseudo-elements, so it shouldn't move.
  Florian: We mean to attempt it.
  dauwhe: I hope it succeeds. I'm so used to it for formatters.

  astearns: So the argument that what I put in the pseudo draft and
            should it go in content as well.
  dauwhe: Yes.

  dbaron: One of the big issues is how it effects resource loading
          and performance of getting them started quickly because
          one of the things authors are concerned about is getting
          the important things moving along the network quickly. If
          you have to load all the CSS before you retrieve images,
          it's a big obstacle.
  tantek: +1 to dbaron and the consequence is that authors either
          won't use the method or they'll use an alternative no
          matter how hacky. If you want this to work it must give
          them the sort of control dbaron is talking about.
  astearns: And I think there would be a big content-style
            separation argument.
  dbaron: Many of the use cases were for having CSS describe a
          markup language, not actual things authors would use.
          We're less interested in CSS being a complete description
          of how HTML works and part of that is the complexity of
          resource loading.

  glazou: Should we continue this or move on to something else for
          four minutes?
  dauwhe: I'm not sure we should have the last five minutes on
          something big.
  Florian: One of mine is short.

:hover pseudo
-------------

  <Florian> http://www.w3.org/mid/FD5224A3-309C-4B65-BD71-D50D4F0314BC@rivoal.net
  <tantek> +1 agreed with Florian's proposal
  Florian: There's a non-normative statement that we should fix.
           Regarding the :hover pseudo when the label is hovered the
           control is also in the hover for HTML
  fantasai: Yeah, we need to change that.
  glazou: So there's an issue and it will be fixed.

Where does ::selection inherit from?
------------------------------------

  <fantasai> http://lists.w3.org/Archives/Public/www-style/2014Nov/0188.html
  fantasai: I've got a topic for ::selection.
  fantasai: I ran some tests on where ::selection inherits from.
            1 is from parent ::selection, 2 are from originating
            element, and 1 where it's different for BG and color.
  fantasai: I wanted to ask implementors which behavior they prefer
            so we can spec it.
  <dbaron> I can't even load this issue into my head in 2 minutes.

  fantasai: We don't have to answer now, but I'd like an answer at
            some point.
  glazou: We should do that on the ML.
  fantasai: The discussion is here (above). I'm not expecting an
            answer now.

  glazou: Thank you. Anything else in the last minute?
  glazou: Okay. Sorry the first issue took so long. Thank you and
          talk to you next week.

Received on Thursday, 13 November 2014 01:10:00 UTC